Ganging Up on Big Data Computer-Intermediated Collaborative Analysis Mark Stefik and Hoda Eldardiry Intelligent Systems Laboratory Palo Alto Research Center (PARC) Palo Alto, California, USA {mark.stefik,hoda.eldardiry}@parc.com Abstract—Understanding complex situations is difficult. In- telligence analysis has long been the work of teams including subject matter specialists. Today collaborative analysis takes place in the context of “big data”, where information comes from a variety of human, communications, and sensor sources. Understanding the big picture is both about how analysts interact and combine their insights together and with how they engage with data at scale. In this paper we consider opportunities for next generation analysis systems for teams, focusing on the computer-intermediated functions that support and coordinate analytic activities around big data. Keywords—Planning; generalization; lessons learned knowl- edge management; collaborative analytics; anomaly detection Figure 1. Activity Analytic Map Citations Teams I. INTRODUCTION Figure 1 shows a descriptive visual analytic from one of our projects. This dashboard highlights activities of a traffic and parking enforcement organization; and provides insights to the organization on its own activities and their interactions with an urban smart city environment. Beyond such descriptive analytics, the opportunity and challenge for impact requires analyzing the work inside individual and team activities. We describe three central functions for sustainable impact in many next generation collaborative analytics settings: • Connecting activities. People engage in different activi- ties, bringing different knowledge and expertise. How can a system leverage big data and this diversity to coordinate and amplify their performance? • Automating Tasks. People working with analytics and big data always know salient things not yet represented in the system. How do we engage them together with automatic processing to improve performance by guiding foraging, monitoring and interpretation of big data? • Generalizing learning. Situations evolve, leaving traces in collected big data. How can lessons from the past be updated to keep up with the emerging future? Figure 2. Framework Template